This project proposes to collect and analyze probabilistic expectations data and other subjective data in an effort to better understand perceptions of economic insecurity, lifetime income expectations, and behavioral responses to changes in public policy. This work should also advance the state of the art of expectations elicitation. Economic Insecurity: The data to be collected will be used to assess the distribution of perceived risks of economic misfortune across households and over time. It will analyze the relationship perceived and realized risks, as well as the relationship between perceived risks and attitudes toward risk. It will also study the behaviors in which households engage in order to avoid and to insure against economic risks. Lifetime Income Expectations: Income expectations are central to economic models of household behavior. This project shall elicit subjective expectations of near-term and long-term house income as well as perceptions of the co-variation of these outcomes. The data will be used to estimate household-specific subjective probability distributions of lifetime income. These estimates will be compared with ones based on realizations data. Elicitation Design: The success of the proposed research hinges on the ability to collect valid, probabilistic expectations data. Such data provide a promising alternative to the conventional approaches of inferring expectations from (1) data on realizations and assumptions about the process of expectations formation or (2) responses to loosely-worded expectations questions. The project will devote considerable attention to testing and refining methods for eliciting probabilistic expectations. The project builds on experience eliciting expectations in the Health and Retirement Study and in the Survey of Economic Expectations (SEE), an ongoing national telephone survey initiated by the investigators. The two surveys are complementary vehicles for data collection. The SEE enables inexpensive and flexible collection of limited data. It is well-suited for development and evaluation of experimental survey methods. The HRS offers the opportunity to collect much richer data, but it requires a longer planning horizon and entails higher costs. The project aims to use in- person interviews and SEE interviews to evaluate questions in the HRS and to develop new questions.